In this study, successful postsecondary completion attained by students who are deaf or hard of hearing is investigated by level of degree using data from the National Longitudinal Transition Study 2 (NLTS2). The NLTS2 is a large-scale national policy, youth-focused database. The purpose of this investigation is to perform an exploratory, retrospective study of the
longitudinal effects, direct and indirect, of student characteristics on the postsecondary educational attainment of students who are deaf or hard of hearing. Youth who attended a postsecondary program and completed a degree were the focus of this study. Those who withdrew before degree completion were not included in this study. A goal of this study is to identify factors that influence postsecondary completion among this population in order to develop appropriate areas of intervention and support that may assist educators, counselors and other practitioners when serving students as they transition from high school to college.
The statistical model used in this project employed multiple methods to answer the questions in this study. IRB approval was granted from the University of Arkansas Fayetteville (See Appendix A) and from the University of Arkansas at Little Rock (See Appendix B). In addition the University of Arkansas at Little Rock requested and received the NLTS2 data from the National Center for Education Statistics and is an approved site for federal data (See
Appendix C). This chapter explains the background of the National Longitudinal Transition Study 2 database, the dependent variables, and independent variables being utilized in this study. This chapter will conclude with a description of the data analysis that was employed using the NLTS2.
Research Questions
This study seeks to investigate factors that contribute to a successful postsecondary completion among students who are deaf or hard of hearing. Explanation of each variable being used can be found in Table 1 (See Appendix E). The following research questions are addressed: 1. What are the personal characteristics of postsecondary graduates who are deaf or hard of
hearing?
• Demographic/student variables: gender, race, cochlear implant user, communication modality, and type of sign language used.
2. What are the family characteristics of postsecondary graduates who are deaf or hard of hearing?
• Family variables: parental academic expectation and parental educational postsecondary attainment.
3. What are the psychological characteristics of postsecondary graduates who are deaf or hard of hearing?
• Psychological variables: self-concept, self-determination, self-advocacy and friendship interaction
4. What are the educational setting and attainment characteristics of postsecondary graduates who are deaf or hard of hearing?
• Educational variables: math score, reading score and high school academic setting (public, private or residential)
5. Among students who are deaf or hard of hearing, what are the relationships of demographic/student, family, psychological, and school related factors to type of
postsecondary completion, specifically vocational/technical certificate, 2-year degree and 4- year degree?
Research Hypotheses
The underlying hypotheses for these research questions are:
1. Demographic and student variables (gender, race, cochlear implant user, communication modality, and type of sign language used) among students who are deaf or hard of hearing strongly influence academic postsecondary completion. The null hypothesis is that
demographic and student variables (gender, race, cochlear implant user, communication modality, and type of sign language used) among students who are deaf or hard of hearing do not influence academic postsecondary completion.
2. Family variables (parental expectation and parental educational attainment) among students who are deaf or hard of hearing strongly influence academic postsecondary completion. The null hypothesis is that family variables (parental expectation and parental educational attainment) among students who are deaf or hard of hearing do not influence academic postsecondary completion.
3. Psychological variables (self-concept, self-determination, self-advocacy, and friendship interaction) among students who are deaf or hard of hearing strongly influence academic postsecondary completion. The null hypothesis is that psychological variables (self-concept, self-determination, self-advocacy, and friendship interaction) among students who are deaf or hard of hearing do not influence academic postsecondary completion.
4. Educational variables (math score, reading score and academic high school setting) among students who are deaf or hard of hearing strongly influence academic postsecondary
academic high school setting) among students who are deaf or hard of hearing do not influence academic postsecondary completion.
5. Student, family, psychological, and educational variables will have a strong relationship with postsecondary educational attainment. The null hypothesis is that student, family,
psychological, and educational variables do not impact postsecondary educational attainment.
Research Design
This research utilized the NLTS2 database to investigate the research questions proposed in this study. The NLTS2 is a large-scale national policy study initially funded in 2000 by The Office of Special Education Programs (OSEP) of the U.S. Department of Education to examine characteristics, experiences and transition outcomes of students with disabilities. Areas of focus included academic performance, high school coursework, extracurricular activities,
postsecondary education and training, community participation, independent living and
employment. The sample in this study reflects all segments of the U.S. population, representing more than 1.8 million students with a disability (Wagner, 2003).
The goal of the NLTS2 study was to have a sample of students that would represent the population of Local Education Agencies (LEAs). Across the United States there are
approximately 12,000 LEAs, so in order to increase precision of estimates, the NLTS2 LEA sample was stratified to ensure low-frequency types of LEAs were adequately represented. The stratifying variables used were region, LEA size, and LEA community wealth. The region variable captures differences among the organizations, political views and economic conditions under which the schools operate.
The regions represented four geographical areas: Northeast, Southeast, Midwest and West. The size of each LEA was sorted into four categories very large, large, medium and small.
A very large LEA had enrollments greater than 14,931 in grades 7 through 12. A large LEA had enrollments from 4,661 to 14,930 in grades 7 through 12. A medium LEA had enrollments from 1,622 to 4,660 in grades 7 through 12. A small LEA had enrollments from 11 to 1,621 in grades 7 through 12. The LEA/community wealth variable was measured using the Orshansky index to ensure the sample of students would be representative of the portion of students living below the level of poverty. The Orshansky index calculates the proportion of the student population living below the federal definition of poverty (Fisher, 1992). The variable was organized into four categories, high (0 percent to 13 percent Orshansky), Medium (14 percent to 24 percent Orshansky), Low (25 percent to 43 percent Orshansky, and very low (more than 43 percent Orshansky).
Once the LEAs were randomly stratified, a sample of 3,635 LEAs and 77 state-supported schools for the deaf and schools for the blind were invited to participate in the study. The initial sample is a random selection of approximately 12,000 students that includes approximately 1,250 students in the disability category. A smaller sample of students was selected among three low-incidence disability categories, i.e., students who are autistic (n=1,012), who have traumatic brain injuries (n=559) and who are deaf-blind (n=122). Of the initial sample surveyed, 501 school districts and 38 special schools participated, which resulted in a sample size of 11,272 students.
Students selected to participate in the sample were from all regions of the nation, including all 50 states and the District of Columbia. The sample represented urban, suburban, and rural regions, and racial/ethnicity groups were equally represented with the categories of Caucasian, African American, Hispanic, Asian/Pacific Islander, and American Indian/Alaska Native. Disability populations were equally represented in this study and were categorized as
students with a learning disability, speech impairment, mental retardation, emotional disturbance, hearing impairment, visual impairment, orthopedic impairment, other health impairment, autism, traumatic brain injury, multiple disabilities, and deaf-blindness. Gender was also equally
represented with 50.8% of the sample being male, and 49.2% were female. The nationally representative sample of youth receiving special education services were between the ages of 13 and 16 years old on December 1, 2000, and were in at least 7th grade.
The data were collected over a nine-year span concluding with data collection activities in 2010. Data collection for the National Longitudinal Transition Study 2 (NLTS2) began in 2001, and is reported according to waves Wave One collected data during 2001-2002, Wave Two collected data during 2002-2004, Wave Three collected data during 2004-2006, Wave Four collected data during 2006-2008 and Wave Five collected data during 2008-2010 (Wagner, Newman, Cameto, Levine and Garza, 2006). A variety of collection strategies was used during this study throughout waves one through five. Parent interviews were conducted during all wave periods. Student interviews and assessments were gathered during waves two through five. Staff interviews were conducted in waves one and two. To date all waves of the project have been completed. A projected timeline and description of the types of data collected is presented in Figure 1 below.
Figure 1. Timeline of milestones in NTLS2 Year 1 2000- 2001 Year 2 2001- 2002 Year 3 2002- 2003 Year 4 2003- 2004 Year 5 2004- 2005 Year 6 2005- 2006 Year 7 2006- 2007 Year 8 2007- 2008 Year 9 2008- 2009 Year1 0 2009- 2010 Parent telephone interviews ★ ★ ★ ★ ★ Youth telephone interviews ★ ★ ★ ★ ★ Direct assessment/ student in- person interviews ★ ★ Teacher survey ★ ★ School program survey ★ ★ School background survey ★ ★ Transcripts ★ ★ ★ ★ Analysis ★ ★ ★ ★ ★ ★ ★ ★ ★ ★
The intent of this study was to analyze waves one through five of the NLTS2. However, the first several copies of wave five that were received from NCES were corrupted and a clean copy of the data was not available at the time of this study. Therefore, only waves one, two, three and four were analyzed. This study utilized data from waves one through four of the National Longitudinal Transition Study 2 data collection: Parent interviews from waves one through four, youth telephone interviews from waves one through four, student assessments from wave one and two, school program surveys from wave one and two, and school background surveys from wave one and two. Independent variables are from waves one, two, and three of the data
collection. The dependent variable, postsecondary completion, is from the fourth follow-up wave. This large-scale national policy, youth-focused database provides the unique opportunity to study how demographic, family, behavioral, psychological, and school experiences influence
postsecondary educational attainment of youth with a hearing impairments.
Participants
For the purpose of this study, only participants in the NLTS2 database identified as “hearing impaired” who enrolled and completed a postsecondary program were analyzed. A postsecondary program is defined as a vocational training facility, technical school, community college or university. The NLTS2 refers to individuals who are deaf or hard of hearing as those with a “hearing impairment” so in order to provide consistency throughout this chapter, the language used to describe this population is congruous with terminology used in the national study. The NLTS2 used the definition that was described in the Individuals with Disabilities Act (IDEA) to describe the population of individuals with a disability. The definition used in the IDEA to describe individuals who are deaf or hard of hearing is:
“An hearing impairment, including deafness is an impairment in hearing, whether permanent or fluctuating, that adversely affects a child's educational performance.
Deafness is a hearing impairment so severe that the child cannot understand what is being said even with a hearing aid” (Knoblauch, 1998).
According to the NLTS2, it is estimated that students with a hearing impairment make up 1.2% of the population, which equals a population size of 22,001 students. Students with a hearing impairment who participated in the NLTS2 study were 51.5% male and 48.5% female. For this study, students with a hearing impairment constituted a heterogeneous population that varied in communication mode, academic placement, type of hearing loss, degree of loss, and age of onset. The racial/ethnic composition was approximately 59.9% “Caucasian,” 17.5% “ African American,” 17.3% “Hispanic,” 4.1% “Asian/Pacific Islander,” and 1.2% “American Indian/Alaska Native.” There were a disproportionate number of Hispanic students with a hearing impairment in comparison to Hispanic youth without disabilities (12.8%) in the general population.
Sampling Weights and Design Effects
The data used in the NLTS2 correspond with many different disability populations. The NLTS2 designed the samples to be used as weighted samples. Analysis weights can be used to identify group populations. A weight is added to the sample in order to adjust for the effects of non-response and is used to compensate for unequal probabilities of selection. The NLTS2 uses weights so that appropriate generalizations about the national population can be made. The purpose of the weights is so that each disability is representative of the national population. If weights were not used in a simple random sample, low-incidence populations would not be adequately represented.
Weights were calculated for each instrument used during the data collection and for each wave period. For specific analytic purposes special weights can be created. Weights should be
used when combining data from two different instruments or waves. For the purposes of this study, weights are applied to the analysis using the weight from the instrument with the lowest number of respondents, because this weight more closely approximates a special cross-
instrument weight.
Assessments
Students in the sample received two types of assessments, a questionnaire and several direct assessments. The questionnaire that students received was administered through telephone interviews. For students with a hearing impairment the questionnaire was mailed to them and either taken individually or with a sign language interpreter. Several direct assessments were administered to students in wave one through phone or in person. Subsets of the Woodcock- Johnson III Research Edition were utilized to assess reading, math and content comprehension, the Student Self-Concept Scale (SSCS), the Self-Determination Scale (SDS), the Friendship Interaction Interview and the Scale of Independent Behavior Revised (SIBR). This study
included the data collected from the reading and math subsets of the Woodcock-Johnson III, the SSCS, SDS and the Friendship Interaction Interview.
Woodcock-Johnson III - Research Edition
The Woodcock-Johnson Psycho-Educational Battery is used with individuals between the ages of two through ninety years old to assess cognitive and academic achievement. The research edition was created by the Woodcock-Johnson III developers for use in the NLTS2 and is a shorter version of the standard assessment. Three domains of the reading portion of the Woodcock-Johnson III were used to assess students’ synonyms, antonyms and passage comprehension. Two domains of the math portion, the applied problems subset and the calculation subset, were used to assess how students analyzed and solved problems, as well as
performed mathematical calculations. For the purpose of this study, the composite math and reading score are used as indicators of early academic performance.
The Student Self-Concept Scale (SSCS)
The Student Self-Concept Scale (SSCS) is a measure of self-concept that was developed by Gresham, Elliott, and Evans in 1993 and is based on self-efficacy theory and subjective task value. This seventy-two item multidimensional measure assesses both efficacy expectation and outcome expectations of children and adolescents in grades three through twelve. The SSCS has ample evidence of reliability and validity and has been nationally standardized (Gresham, 1995). The SSCS has three content domains: self-image, academic self-perception and social self- perception. Of these domains, scales are reported according to three rating dimensions: self- confidence, importance and outcome confidence. Internal consistency has been documented using Cronbach’s alpha, reliabilities ranged from .89 to .92 across all elementary and secondary students (Gresham, 1995).
The NLTS2 chose 30 items from the SSCS to include on the direct assessment given to youth. Fifteen items from the self-confidence domain and importance domain were chosen to be included in the assessment. Sample items from subscales are described in Figure 2.
Figure 2. Sample Items from SSCS Subscales Self-image
I am easy to like. I am a nice person. I am proud of who am.
Academic
I can do my homework on time. I can read aloud in class without feeling nervous.
I can do my math work without help.
Social
I like to be with others. I can control my temper in arguments with other kids. I can make friends easily. The total average scale score for the SSCS is used as the subjects’ level of self-concept in this study.
The Self-Determination Scale (SDS)
Wehmeyer and colleagues (1995) constructed the Self-Determination Scale (SDS) as a self-report measure of self-determination to be used by youth with disabilities. Self-
determination is defined as, “acting as the primary causal agent in one's life and making choices and decisions regarding one's quality of life free from undue external influence or interference” (Wehmeyer, 1995, p. 1). The scale is designed to assess four characteristics of self-
determination: “(a) the individual acts autonomously; (b) the behaviors are self-regulated; (c) the person initiates and responds to event(s) in a “psychologically empowered” manner; and (d) the person acts in a self-realizing manner (Wehmeyer, 1995).” The scale has strong internal
consistency reliability with a Chronbach alpha of .90 for the scale as a whole.
Rather than administer the scale in its entirety, the NLTS2 selected portions of the Self- Determination Scale (SDS) to administer to youth. The self-regulation subscale was not included in the direct assessment portion of the NLTS2. This section is a cognitive problem-solving approach through story telling. Only 15 out of the 32 questions on the subscale Autonomy were administered to subjects. Questions employed a four point Likert scale; 1=Not when I have the chance, 2=Sometimes, 3=Most of the time, and 4=Every time I have the chance. Sample
questions included, “I keep my personal items together, I keep good personal care and grooming, I make friends with other kids my age, etc.”
The second subscale used in the NLTS2 from the SDS was the self-realization subscale. Only 5 items from this 15-item subscale were incorporated into the direct assessment. Again a four-point Likert scale was used, and response choices included; 1=Never agree, 2=Sometimes agree, 3=usually agree, and 4=Always agree. The five items included on the NLTS2 from the SDS self-realization subscale are; “I can like people even if I don’t agree with them, I know what
I do best, I like myself, I know how to make up for my limitations, and I am confident in my abilities.”
The final subscale used from the SDS was the psychological empowerment section. This domain of the scale prompts subjects to check the answer that best describes them. Six out of sixteen items were chosen to be included in the NLTS2 direct assessment. Sample questions asked the youth to choose whether, “I tell others when I have new or different opinions/ideas or I usually agree with other’s opinion/ideas; I can make my own decisions or other people make decisions for me.” The total average scale score for the SDS is used as the subjects’ level of self- determination in this study.
Friendship Interaction Interview
Friendship interaction was measured by including two items from the 24-item Loneliness portion of the Friendship Interaction questionnaire developed by Asher (1984). Of the 24 items on the questionnaire, 16 are questions related to loneliness and were found to be internally consistent with a Cronbach’s alpha of .90. The two items included on the NLTS2 from the loneliness domain of the Friendship Interaction interview were, “I can find a friend when I need one, and I’m lonely at school.” Responses for these two items include 1=yes, 2=no, and
3=sometimes.
Research Variables Educational Attainment Variable
All of the variables of this study (see Table 1) were measured by information gathered from parent, youth and school questionnaires and interviews or through direct assessments. The purpose of this study is to examine which factors contribute to postsecondary completion. The variables used to verify educational postsecondary completion were taken from parent and youth